v2.1.0

Node binding changes:

• feat: rewrite the node bindings in TypeScript
• feat: add console.warn for bad inputs to the functions

C++ changes:

• fix: get_at wasn't working for std::vector<std::string>
• docs: add usage from C++ docs

v2.0.2

Include the entry file in the npm package.
See: https://github.com/atom-community/zadeh/releases/tag/v2.0.0

v2.0.1

Include zadeh.d.ts in the npm package.
See: https://github.com/atom-community/zadeh/releases/tag/v2.0.0

Rename the library to Zadeh + Add direct C++ interface

• I have changed the library's name to Zadeh as a reference to Zadeh, the inventor of fuzzy logic.

• Add direct C++ interface:
  This library should be usable directly from C++

  • Changed the folder structure.
  • Make the files header-only
  • Add an index file that is callable from C++ (it should include the implementations from binding/node.cc
  • Adds ArrayFilterer and TreeFilterer class that are callable from C++. These are used in Node bindings

Breaking Changes

This deprecates the old npm package (no more updates). The new package will be published starting from version 2.0.0:
The API has not changed, you just need to install the new package, and rename the library in the places you used.

v1.2.4

This release has several performance optimizations.

• The most important one is in the internal filter algorithm, which now uses vector instead of priority_queue which gives it around 30% speedup in some benchmarks
• Now, the number of threads used in the filter algorithm is automatically detected based on the actual threads of the machine.

The benchmark scripts are also improved to use better measurements and printing.

v1.2.3

• TreeFilterer minor optimization
• Make sure that the prebuilds are built on the latest operating systems
• Make sure that prebuilds work on Electron 9

1.2.2

Several C++ optimizations:
v1.2.1...1.2.2

v1.2.1

C++ Optimizations and exception safety fixes:
v1.2.0...v1.2.1

v1.2.0

This release allows setting the candidates only once using ArrayFilterer and TreeFilterer classes, and then, perform filter multiple times. This is much more efficient than calling the filter or filterTree functions directly every time.

See readme for more information and examples:
https://github.com/atom-ide-community/fuzzaldrin-plus-fast#arrayfilterer-class
https://github.com/atom-ide-community/fuzzaldrin-plus-fast#treefilterer-class

#42

v1.1.13

Ensure that prebuilds are built with release config.